Radiation damage of the silicon detectors in future hadron colliders poses a major challenge for its reliable operation. It is crucial to investigate the neutron-induced radiation damage owing to its large flux in the calorimeters and the trackers. Measurement results on irradiated detectors are complemented by modeling for getting a deep insight into the device behavior. This work presents the development of neutron-induced radiation damage model in silicon detectors using TCAD simulation. A wide spectrum of measurement results available on neutron-irradiated silicon detectors, viz. full depletion voltage, leakage current, charge collection, and effective trapping times, helps in constraining the model parameters, resulting in a robust model for neutron damage. Modeling has been performed within the phase-space of the measurements, i.e. for different initial bulk resistivities, active thicknesses and for both polarities up to the fluence values of 9×10 14 1 MeV n eq cm −2 (where, 1 MeV n eq refers to the equivalent fluence for monoenergetic neutrons of energy 1 MeV) and for two measurement temperatures, 253 and 263 K. The results obtained from the devised neutron damage model show a good agreement with the measurement results.

The low gain avalanche detector (LGAD), having a unique feature of built-in charge multiplication, is more efficient in terms of charge collection (CC) than the traditional silicon detector even after irradiation. However, a dramatic decrease in the charge multiplication beyond a fluence of 3 × 1014 n eq ⋅ cm−2 is observed in the measurements. In the reported work, TCAD CC simulations are carried out on various physical and geometrical LGAD design parameters with the aim to understand and extend the radiation hardness capabilities. It is observed that a thin LGAD with low bulk resistivity may survive up to a fluence of 3 × 1015 n eq ⋅ cm−2 for an optimal choice of p-well design. A detailed investigation including CC and leakage current validation with experimental data and 1D electric field profile, in support of optimizations performed, is also provided.